I’m not an expert on Google Analytics, though I do use them on my sites and monitor them regularly. But there is a huge caveat on using analytics in tech comm which a Google Analytics expert from another field may not be aware of.
We want analytics to determine the quality of content and its value. The problem is, visitor statistics tell you very little about the quality or the value of technical communication.
For regular content, quality content gets more visits. But for tech comm, the number of visits to a particular topic is determined primarily by the number of times that the user encounters the problem that that topic addresses. This has nothing to do with the quality of the topic. It is all about the quality of the product and the things people do with it.
Now we might think that if the analytics don’t tell us a lot about quality, they do at least tell us a lot about value, that the more often a topic is read, the more value it is delivering to the customer. But this is not necessarily so either.
Some user problems are more expensive/valuabe than others. A problem like the main server crashing and corrupting the main database could be a problem that costs a million dollars an hour. But if your server crashes and corrupts your database on a regular basis, your product is going to be off the market real quick. If your product is reliable enough to be commercially viable, the topic on recovering from a server crash that trashes the database is probably the single most valuable topic in your entire documentation set, and probably the one least often read.
Google analytics is not going to tell you that.
This is not to say there is no value to be had from Google analytics, but it is important to remember that the relationship between quality, value, and page views for tech comm is radically different than it is for content marketing, for instance.